It's been a tough few weeks for gold investors. In mid-July this Ancient Metal of Kings closed near $976, within spitting distance of its all-time nominal high of $1005 achieved in mid-March 2008. But since then it has sold off relentlessly, down 10.4% ($101) at worst. Several days of this ugly span were marred by sharp selloffs too, including this past Tuesday's 2.2% plunge on the Fed's machinations.

If you are playing gold on the long side, it is tempting to get discouraged. After all, this week CNBC was emphatically claiming that the global commodities bulls had ended. Never mind that CNBC has prematurely declared these bulls over about a dozen times in the past few years, the misery of the moment is hard to transcend. Gold certainly wouldn't fare well if the commodities-hating financial media is right.

Thankfully they are not. Global supply deficits in commodities that took decades of infrastructure neglect to create can't be magically fixed overnight. It will take decades to find mineral deposits, get permits, sink the mines, and bring these scarce natural resources to market to satiate soaring world demand. Gold is no exception, as it is often considered one of the most challenging and difficult metals to mine.

Provocatively gold's weakness of late was expected by students of the markets. It was nothing unusual at all. Gold has strong seasonal tendencies in its price behavior and the summers are its weakest time of the year statistically. Just a couple weeks ago I wrote an essay on the PM Summer Doldrums that delved into this period of time specifically. Gold seldom does anything but grind sideways during the summer months.

Many traders are surprised when they learn seasonality affects gold. It makes sense for the soft commodities like wheat where planting and harvest, hence relative scarcity and abundance, are slaved to the Earth's celestial mechanics that drive our calendar seasons. In their case, most new supplies come to market during a relatively tight calendar-defined window at harvest. This affects prices, of course.

But seasonality can even affect commodities that are produced fairly uniformly year round. In these cases it is not supply that fluctuates seasonally, but demand. If demand tends to cluster around certain parts of the calendar year, definite price impacts will be observed. For example snow shovels can be manufactured anytime, but demand (and price if hardware stores are good capitalists) only spikes after major blizzards.

In gold's case, its seasonality is governed exclusively by investment demand. This is always the biggest wildcard affecting the gold price. There are times of the year when investors rush to buy gold in various parts of the world, and times of the year when investors nearly forget about gold. These episodes of varying gold-demand intensity drive major price fluctuations that hit at specific times of the calendar year.

Before we dig into this gold seasonality, an important caveat is in order. Seasonal influences are real and tradable, but they are often just a secondary driver. Technical extremes that spawn sentiment extremes (excessive greed or fear) can easily override seasonals. I think of seasonals like prevailing winds. While you don't need a tailwind to drive your car down the highway, it is certainly nice to have.

In order to chart gold's seasonal trends within this secular bull, I individually indexed every year between 2000 and 2008. The first close in January was assigned a value of 100, and then the rest of the year's closes were indexed off of that. Indexing makes every year perfectly comparable in percentage terms regardless of gold's rising general price levels endemic to this bull.

Next I averaged each individual-year index to arrive at the blue gold seasonals line below. It reveals what the gold price has tended to do at any point in the calendar year since 2000. Specific days are rendered with large dots in order to clearly show where the actual data is relative to the lines that connect these datapoints. I also included standard-deviation bands to show how dispersed the underlying data was.

Averaging 10 and 90 or 48 and 52 both yield 50. But as a speculator the second tighter data set makes me much more comfortable about the odds a tendency will repeat itself. So the tighter the yellow standard-deviation bands, the tighter the underlying pre-average data. Narrower bands around any seasonal time mean that particular seasonal tendency is less likely to be the product of anomalous extremes and more likely to repeat itself.

The first conclusion your eyes reap from this chart is probably gold's tendency to rise on balance. Throughout the calendar year, it is generally climbing. This is logical since these seasonal tendencies only apply to the years since 2000 where gold has been in a strong secular bull. While bulls do flow and ebb, advancing then correcting, over time they inevitably lead to higher price levels.

Seasonally gold has carved a definite uptrend channel for the first 2/3rds of the calendar year. It is really well-defined too, with four solid intercepts at support and three at resistance. Like any conventional technical uptrend, traders can game this tendency. The seasonal tailwinds are more likely to blow gold higher when it is near support and lower when it is near resistance.

Regardless of what primary trading indicators you use, it is still wise to consider prevailing seasonals before you launch new trades. Ideally your primary and secondary indicators all line up, which gives you better odds for success in your trades. And if your primaries don't agree with seasonals, the primaries will probably still prove right yet the resulting move isn't likely to be as intense since it has to buck seasonal headwinds.

Within this seasonal uptrend, the biggest rally to exploit runs from mid-March to late May. Averaged across all the years since 2000, gold tends to run higher over this short span. Unfortunately this seasonal tendency failed in 2008, as you certainly remember if you were long gold then. Between mid-March and late May gold fell about 12% this year. Yet even with this bad 2008 averaged in above, gold's seasonal tendency is still to run higher over this particular span.

The events of spring 2008 are a great cautionary tale on reading too much into seasonality. Excessive greed or fear, regardless of their cause, can overwhelm seasonal tendencies. So once again make sure you use seasonals only as a secondary trading indicator. While they do help define the probabilities of a price moving in a certain direction, the probability scale still leaves plenty of room for variation.

After the usual spring gold rally, summer is this metal's weakest time of the year by far. Note above how gold merely tends to grind sideways on balance in June, July, and August. I just wrote an essay on this tendency alone if you want to dig deeper. The best way to weather gold's summer doldrums is to not expect too much from the metal and realize it is likely to simply trade sideways in a range-bound fashion.

But boy, if you can weather the summer doldrums without psychological damage, your prize is autumn. Autumn is the strongest time of the year for gold seasonally by far. From early August (like right now!) to early February, on average gold has rallied 14.0% in its bull to date. This is a big move, the part of the year in which most of gold's bull-market gains have been achieved. Be long gold between August and February!

This massive autumn rally starts accelerating in late August and gold blows above its seasonal resistance by mid-September or so on average. While going long in August is optimal, there is a brief pullback in early October that offers procrastinators one last chance to get long. And after that gold just powers higher without material respite into early February. Gold's seasonal strength over this period is awesome.

But why? What makes August to February so special? Since new gold brought to market is relatively constant throughout the year, it has to be a demand phenomenon that drives this big autumn rally. Investment demand for gold has to be much more intense between August and February than it is for the rest of the year. And this is indeed the case as a variety of cultural factors drive a surge in gold buying.

Starting in late August or early September, Asian crops are harvested. The vast majority of Asia (all of continental Asia) sits in the northern hemisphere, so it enjoys the same sidereal seasons as we do in the States. Thus Asian farmers harvest their crops at the same time of the year we do. But unlike American farmers, Asians tend to plow some of the surplus fruits of their labors into physical gold bullion.

Gold is the ultimate form of saving. It has outlived every failed government and fiat currency regime in history and still retains its intrinsic value to this day. It will outlive every government and currency on the planet today too. So if you want to protect your surplus labor, save it to build wealth, gold is the best option. And this is especially true in countries with lower political stability than we enjoy, which is virtually all of Asia.

The demand spike in gold in late August and September that drives sharply higher prices most years is a result of this harvest buying. Crops are sold for cash, and some portion of this cash not directly needed for ongoing expenses is converted into physical-gold savings. It is too bad Americans are not smart enough to do this, both to save (consume less than we earn) and to store wealth outside of the ailing US dollar in gold.

The wise and prudent Asians, having lived through countless failed governments and currencies over millennia, have a deep cultural affinity for gold. This continues after harvest into the famous Indian wedding season. This fascinating cultural phenomenon tends to peak between early October and late November, and is directly responsible for those months' strong gold rallies.

In India, weddings are a huge deal. Most marriages are arranged, and couples are typically married off during autumn festivals like Diwali. It is believed that being married in festival season provides good luck, longevity, happiness, and success for a marriage. The families of Indian brides give them wedding gold in the form of intricate 22-karat jewelry. Not only is it beautiful adornment for the bride, but gold's intrinsic value helps secure her financial future and her financial independence within her husband's family.

India is the world's largest consumer of gold. Most of it is in the form of jewelry, but Indians don't separate gold jewelry and gold investment like we do in the West. They are one and the same. Brides' dowries may not sound like much, but collectively they are the biggest seasonal driver of gold investment demand on the planet. Something like 40% of India's entire annual gold demand occurs during the short autumn wedding season!

Sometimes Westerners marvel at this, yet we aren't all that different. As the tail end of Indian wedding season arrives, gold demand surges in the West for Christmas buying. A big portion, if not the majority, of discretionary spending in the West occurs between Thanksgiving and Christmas. Some of these holiday dollars flow into gold jewelry as gifts for wives, girlfriends, daughters, and mothers. So Western gold jewelry demand is also concentrated seasonally into a narrow period of time, essentially December.

After Asian harvest, Indian wedding season, and Western holiday buying run their courses, you'd think that gold investment demand would wane dramatically. But this isn't the case. One more event spikes global investment demand, Chinese New Year. The Chinese calendar is heavily influenced by lunar cycles, so its new year tends to occur between late January and mid-February on our Western calendar.

Gold is woven throughout the Chinese New Year mythos. For example, one of the popular icons for the celebrations is yuanbao. This symbolizes money and wealth and is shaped in the form of ingots that were the standard medium of exchange in ancient China. Gold trinkets are used to decorate homes for festivities and are also given as gifts. This drives strong Chinese gold demand.

I also suspect the Chinese, late in their year (January for us), tend to invest some of their surplus capital in gold. In the West we make many investment decisions late in our own year too, as that is when we finally know how much we made, how much we owe in taxes, and how much surplus capital we can save and invest. Chinese New Year celebrations and Chinese year-end gold buying keep gold buoyant into early February.

So as you can see, gold's strong seasonal rally between August and February is quite logical. Over this span various cultural practices combine to create one long investment-demand-driven surge for the yellow metal. Within these months at various times, all of gold's major consumers have a big cultural reason to buy. And much of this demand isn't economically sensitive. If you are an Indian father marrying off a beloved daughter, I bet you really don't care whether business was good that year or not. You will buy her gold.

Traders would do well to be long gold, and anything PM-related, for this August-to-February period. Almost all of gold's bull-market gains have been made within this investment-demand-intensive window. The optimal timing to get long is psychologically challenging though. Investors and speculators need to be aggressively adding gold positions in August at the dismal demoralizing lows of the summer doldrums. They have to force themselves to buy when they least want to, to be true contrarians.

This next chart looks at gold bull seasonals in a different way, indexed monthly. While slicing the gold price up in calendar months is somewhat arbitrary (trends seldom begin or end on the 1st or 31st), it is still interesting. This approach shows which calendar months tend to be the strongest for gold. Not surprisingly, all of the best occur within the big-autumn-rally span of time.

In order to be strong, gold has to approach a 2% gain in a given calendar month on average over the years since 2000. And I define weak months as a 1% loss on average. This asymmetry exists because we are in a secular bull where prices are generally rising. The same five optimal seasonal times to go long gold rendered in the first chart are replicated here for comparability. Early August is the best of them all.

This is because gold's biggest calendar month of the year on average in this bull has been September. Gold tends to rise by 3.5% that month. This might not seem impressive, but it really is for gold. If 150,000 tonnes of gold have been mined in world history, then the global above-ground gold is worth something like $4.3 trillion at $900 per ounce. So even a 1% move in gold represents several tens of billions in wealth creation or destruction. And of course these seasonal numbers are averages, which moderate the underlying results.

After gold's best month in September, it tends to consolidate into early October in the couple-week lull between the Asian harvest and the Indian wedding season. But then in late October it starts surging and this continues into November. At a 2.5% monthly gain on average, it is gold's second-best month of the year seasonally. Number three follows right after that, with December's 2.2% average gain. And then January comes in fourth at 1.9%.

So of the six months between early August and early February, gold's massive seasonal autumn rally, fully four are gold's biggest months of the calendar year. You absolutely want to be long gold, and indeed the entire PM-complex since everything PM-related ultimately follows gold's lead, in September, November, December, and January. Seasonal-demand-driven price increases are very compelling then.

Obviously this is really exciting today since we are now on the verge of gold's biggest seasonal rally of the year. But remember that seasonals are a tailwind, a secondary indicator. So if gold was overbought today and greed abounded, the bullish seasonals could easily be overridden. But thankfully it is not, indeed just the opposite has occurred. Gold is deeply oversold today and sentiment is horrendous. Excessive levels of both fear and frustration have conspired to create an explosively-bullish sentiment mixture.

I've been discussing these bullish gold, silver, and PM-stock technicals lately in our acclaimed monthly and weekly subscription newsletters. With primary and secondary indicators all lining up and calling for a major gold upleg in the coming months, we've also started adding new PM trading positions. This week's excessive selling has created a very-high-potential buying opportunity that we are exploiting. If you want to see exactly how we ride this thing, and have the chance to mirror our trades, please subscribe today!

The bottom line is gold does have strong seasonal tendencies. Even though gold isn't grown like wheat, the passage of the calendar influences gold investment demand across the globe which directly impacts the gold price. Gold is deeply woven into cultures around the world and their various customs create lumpy gold investment demand. It is clustered at specific times instead of spread out evenly across the year.

Naturally investors and speculators should exploit these seasonal tendencies. The best time seasonally to go long gold and other PM-related trades is right now. From August to February gold's biggest seasonal rally of the year erupts. During this timespan, which includes gold's four best calendar months, the lion's share of its entire bull-market gains have been made. I fully expect the rest of 2008 to unfold according to this precedent.